Confirm Action

Are you sure you want to proceed?

Is minimax-m3-high good at Summarization & Meeting Notes?

minimax-m3-high ranks #74 of 107 for Summarization & Meeting Notes — excellent. The top pick for this task is claude-opus-4.5.

#74 / 107
Rank for this task
92.8
Score
$0.0161
Cost / run

minimax-m3-high on each Summarization & Meeting Notes sub-task

Faithfulness Under Pressure 100.0/100 #1
Transcript Q&A 100.0/100 #1
Action-Item Extraction 95.3/100 #45
Executive Summary 85.8/100 #85
Messy Transcript 74.0/100 #98

Real examples, graded

WinNorthwind pipeline review 100/100

“The summary perfectly captures all outcomes, corrections, and unassigned tasks without any hallucinations or distortions. It correctly notes the $24k deal size, the shelved 8% price increase, Dana's action item for Thursday, Priya's uncommitted report, and the unassigned dataset cleanup.”

WinFerrovia Q2 business review 100/100

“The summary is perfectly faithful to the transcript. It accurately captures the SOC 2 blocker and action item, explicitly notes that the packaging decision is deferred without assuming consensus to bundle, and correctly distinguishes the Meridian new deal from the renewal. The structure is clean, concise, and outcome-focused.”

WinNorthwind — owners, dates, and a no-owner task 100/100

“The summary is flawlessly faithful to the transcript. It perfectly captures the action items, correctly attributes them, includes the exact due dates where stated, and explicitly preserves the caveats around Priya's tentative report and the unassigned dataset cleanup. It also correctly notes the withdrawn price increase without turning it into an action item.”

← Full minimax-m3-high review All Summarization & Meeting Notes rankings → Top pick: claude-opus-4.5 →

Frequently asked

Is minimax-m3-high good at Summarization & Meeting Notes?

minimax-m3-high ranks #74 of 107 models we tested for Summarization & Meeting Notes, scoring excellent.

What is minimax-m3-high's strongest Summarization & Meeting Notes skill?

Its best sub-task here is Faithfulness Under Pressure.

This page is Spring Prompt, running

We just did this for every model. Do it for your prompt.

The rankings above come from running real tasks through real models and scoring every output. Spring Prompt is that same engine — pointed at your prompt, your test cases, and your definition of good.

  • Generate test cases from your prompt — no eval set required to start.
  • Compare models side by side with quality, cost and latency in one matrix.
  • Optimise the winner until the scores say it's ready to ship.
Experiment · Cold outreach email

Prompt × model results

12 test cases · 3 evals
Claude Opus
GPT-5
Gemini
v1
7.1
6.8
7.4
v2
8.3
7.9
8.0
v3
9.2
8.6
8.4
Best combo: v3 × Claude Opus
9.2 quality · $0.004/run · 1.8s